Research Area:  Wireless Sensor Networks
Wireless sensor networks (WSNs) consist of spatially distributed low power sensor nodes and gateways along with sink to monitor physical or environmental conditions. In cluster-based WSNs, the Cluster Head is treated as the gateway and gateways perform the multiple activities, such as data gathering, aggregation, and transmission etc. Due to improper clustering some sensor nodes and gateways are heavily loaded and dies early. This decreases lifetime of the network. Moreover, sensor nodes and gateways are constrained by energy, processing power and memory. Hence, to design an efficient clustering is a key challenge in WSNs. To solve this problem, in this paper we proposed (1) a clustering algorithm based on the shuffled complex evolution of particle swarm optimization (SCE-PSO) (2) a novel fitness function by considering mean cluster distance, gateways load and number of heavily loaded gateways in the network. The experimental results are compared with other state-of-the-art load balancing approaches, like score based load balancing, node local density load balancing, simple genetic algorithm, novel genetic algorithm. The experimental results shows that the proposed SCE-PSO based clustering algorithm enhanced WSNs lifetime when compared to other load balancing approaches. Also, the proposed SCE-PSO outperformed in terms of load balancing, execution time, energy consumption metrics when compared to other existing methods.
Author(s) Name:  Damodar Reddy Edla, Mahesh Chowdary Kongara and Ramalingaswamy Cheruku
Journal name:  Wireless Networks
Publisher name:  Springer
Volume Information:   volume 25, pages 1067–1081 (2019)
Paper Link:   https://link.springer.com/article/10.1007/s11276-018-1679-2